Find patterns within column and attach values in a new column using Python-CodePudding

I have a dataset where, whenever I see a certain word that contains specific words, I'd like to match specific values to within a new column.

Data

id                     status
see-dd-23aaaa33_00     y
see-dd-aaaaa_o00       y
sal-led-sss_0          y
sal-led-sss.AA         n
dis-dd-red_0           n

Desired

id           status pw  
see-dd-2333     y   14  
see-dd-aaaaa    y   14  
sal-led-sss     y   8   
sal-led-sss     n   8   
dis-dd-red      n   5

Doing

I am thinking I can use a dictionary. Whenever I see a pattern of 'see-dd', I'd like to supply the numerical value of 14. When I see a word that contains 'sal-led' I wish to supply the 8 numerical value. Whenever I see 'dis-dd' I would like to match this with the value of 5.

out= {
    'see-dd': 14,
    'sal-led': 8,
}

Any suggestion is appreciated.

CodePudding user response：

Try with replace

out= {
    'see-dd': 14,
    'sal-led': 8,
    'dis-dd':5
}
df['new'] = df.id.replace(out,regex=True)
df
                   id status  new
0  see-dd-23aaaa33_00      y   14
1    see-dd-aaaaa_o00      y   14
2       sal-led-sss_0      y    8
3      sal-led-sss.AA      n    8
4        dis-dd-red_0      n    5

CodePudding user response：

You can take the below data frame and apply the matcher function based on the pattern dictionary 'out' and 'id' column of the df.

out = {'see-dd': 14, 'sal-led': 8, 'dis-dd': 5}

def matcher(row_data):
    for key, val in out.items():
        if key in row_data:
            return val

#This will create a new column 'pw' using your 'out' patterns and values
df['pw'] = df['id'].apply(matcher)